This invention relates to separation systems and their components and more particularly to separation systems and components involving monolithic permeable polymeric materials.
Monolithic macroporous materials such as for example organic monolithic macroporous polymeric materials and monolithic silica packings are known as components for separation systems such as chromatographic or extraction systems. One class of such materials is formed as a monolithic macroporous polymer plug or solid support produced by polymerizing one or more monomers in a polymerization mixture that includes at least a porogen. It is known for some polymerization mixtures, to include other materials such as cross-linking agents, catalysts and small soluble polymers which can be dissolved after polymerization to control the porosity and pore size distribution. Moreover, the plug may be modified after being formed to add functional groups.
The plug or solid support is normally contained in a housing such as for example a chromatographic column or a pressure vessel. The portion of the housing where the plug resides acts as a reactor. In one prior art process for making monolithic columns, the polymerization mixture may be added to the column casing and polymerization initiated therein to form a macroporous polymeric plug or solid support within the walls of the column.
There are wide applications of these plugs or solid supports including gas, liquid and supercritical fluid chromatography, membrane chromatography and filtration, solid phase extraction, catalytic reactors, solid phase synthesis and others. The efficiency of the column or other container for the plug or solid support, the time required for a separation, and the reproducibility of the columns or other container for the plug or solid support are important commercial factors. The efficiency of separation systems such as chromatographic columns with porous polymer in them is related to both the selectivity of the column or other component containing the macroporous polymeric material and to zone spreading. Some of these factors are affected by molecular diffusion and velocity of the mobile phase in the plug or solid support during a separation process.
The manner in which molecular diffusion and velocity of the mobile phase affects column efficiency can be in part explained by showing the effect of these factors on Height Equivalent to Theoretical Plates (HETP), the conventional designation of column efficiency. The van Deemter Equation shows the relationship between zone spreading, flow velocity and diffusion in terms of H (HETP) as follows:
H=A+B/u+Cu with low H corresponding to high efficiency
U=Flow velocity of the mobile phase
A=Radial Eddy Diffusion coefficient
B=Longitudinal Molecular Diffusion coefficient
C=The mass transfer coefficient
Molecular diffusion depends on the diffusion of the molecules but not on the packing of the bed. Eddy diffusion depends on the homogeneity of the packing of the particles.
Zone spreading from mass transfer can be minimized by using non-porous particles and porous particle with sizes smaller than 1.5 microns. However, packing with non-porous particles has extremely low surface area which is detrimental to the purification process (as opposed to the analytical process) because the purification process requires high sample loading. The use of very small packed particles requires either high pressure which is difficult in most of the separation process using current instrumentation or low velocity. which can increase the time for a given separation (sometimes expressed in H per minute).
The prior art separation systems that include as a component macroporous polymeric monolithic plugs or solid supports use plugs or solid supports formed from particles in the polymers that are larger than desired, less homogenous and include micropores. The large size of the particles and their lack of homogeneity result in a lack of homogeneity in the pore size distribution. The non-homogeneity of the pore sizes and large amount of micropores in the prior art porous polymers contributes greatly to the zone spreading as shown by the van Deemter Equation. The large number of micropores contributes to zone spreading by capturing sample and retaining it for a time. This may be stated conventionally as the non-equilibrium mass transfer in and out of the pores and between the stationary phase and the mobile phase.
The prior art plugs or solid supports formed of porous polymers have lower homogeneity of pore size, less desirable surface features and voids in their outer wall creating by wall effect and thus higher zone spreading and lower efficiency than desired in separation systems.
The prior art also fails to provide an adequate solution to a problem related to shrinkage that occurs during polymerization and shrinkage that occurs after polymerization in some prior art porous polymers. The problem of shrinkage during polymerization occurs because monomers are randomly dispersed in the polymerization solution and the polymers consist of orderly structured monomers. Therefore, the volume of the polymers in most of the polymerization is smaller than the volume of the mixed monomers. The shrinkage happens during the polymerization in all of the above preparation processes. One of the problems with shrinkage after polymerization occurs because of the incompatibility of a highly hydrophilic polymer support with a highly hydrophilic aqueous mobile phase or other highly polar mobile phase such as for example, a solution having less than 5-8 percent organic solvent content.
Shrinkage of the porous polymeric materials used in separation systems and their components during polymerization results in irregular voids on the surface of the porous polymers and irregularity of the pore size inside the polymer, which are detrimental to the column efficiency and the reproducibility of the production process. One reason the column efficiency is reduced by wall effect is that wall effect permits the sample to flow through the wall channels and bypass the separation media. One reason the reproducibility of the production process are reduced by wall effect is the degree of wall effect and location of the wall effect are unpredictable from column to column.
The columns with large channels in the prior art patents cited above have low surface area and capacity. The low capacities of the columns are detrimental for purification process which requires high sample loadings. In spite of much effort, time and expense in trying to solve the problems of shrinkage, the prior art fails to show a solution to reduced capacity.
Because of the above phenomena and/or other deficiencies, the columns prepared by the above methods have several disadvantages, such as for example: (1) they provide columns with little more or less resolution than commercially available columns packed with beads; (2) the separations obtained by these methods have little more or no better resolution and speed than the conventional columns packed with either silica beads or polymer beads, particularly with respect to separation of large molecules; (3) the wide pore size distribution that results from stacking of the irregular particles with various shapes and sizes lowers the column efficiency; (4) the non-homogeneity of the pore sizes resulting from the non-homogeneity of the particle sizes and shapes in the above materials contribute heavily to the zone spreading; (5) the large amount of micropores in the above materials also contributes greatly to the zone spreading; and (6) shrinkage of the material used in the columns reduces the efficiency of the columns. These problems limit their use in high resolution chromatography.
U.S. Pat. No. 5,453,185 proposed a method of reducing the shrinkage by reducing the amount of monomers in the polymerization mixture using insoluble polymer to replace part of the monomers. This reduces the shrinkage but is detrimental to the capacity and retention capacity factor of the columns which require high amount of functional monomers. There is nothing mentioned in these patents regarding the detrimental effect of shrinkage on resolution and the resulting irregular voids on the surface of the porous polymer and irregularity of the pore size inside the polymer, which are detrimental to the column efficiency and the reproducibility of the production process.
Prior art European patent 1,188,736 describes a method of making porous poly(ethylene glycol methacrylate-co-ethylene glycol dimethacrylate) by in situ copolymerization of a monomer, a crosslinking agent, a porogenic solvent and an initiator inside a polytetrafluoroethylene tube sealed at one end and open at the other end. The resulting column was used for gas-liquid chromatography. This prior art approach has the disadvantage of not resulting in materials having the characteristics desirable for the practical uses at least partly because it uses polymerization in a plastic tube with an open end.
U.S. Pat. Nos. 2,889,632, 4,923,610 and 4,952,349 disclose a method of making thin macroporous membranes within a sealed device containing two plates and a separator. In this method the desired membrane support was punched out of a thin layer of porous polymer sheet and modified to have desired functional groups. The layers of porous sheets are held in a support device for “membrane separation”. These patents extended the method described in European patent 1,188,736 to prepare a porous membrane and improve the technique for practical applications in membrane separation. The resulting material is a macroporous membrane including pores from micropores of size less than 2 nanometers to large pores. The size of the particles of the polymer is less than 0.5 micrometers. The separation mechanism of membrane separation is different from that of conventional liquid chromatography.
This porous material has several disadvantages, such as for example: (1) the thinness of the membrane limits its retention factor; and (2) the pores formed by these particles are small and can not be used at high flow rate with liquid chromatography columns that have much longer bed lengths than the individual membrane thicknesses. The micropores and other trapping pores trap molecules that are to be separated and contribute to zone spreading. The term “trapping pore” in this specification means pores that contribute to zone spreading such as pores ranging in size from slightly larger than the molecule being separated to 7 times the diameter of the pore being separated.
U.S. Pat. Nos. 5,334,310; 5,453,185 and 5,728,457 each disclose a method of making macroporous poly(glycidyl methacrylate-co-ethylene glycol dimethacrylate) polystyrene in situ within sealed columns. This method extends the methods described in both the European patent 1,188,736 and U.S. Pat. Nos. 2,889,632, 4,923,610 and 4,952,349 for preparing liquid chromatography columns for the separation of proteins. U.S. Pat. Nos. 5,334,310, 5,453,185 and 5,728,457 profess the intention of improving the column efficiency by removing the interstitial volume of conventional packed columns having beads. The plugs formed according to these patents have a pore size distribution that is controlled by the type and amount of porogens, monomers and polymerization temperature. The macroporous polymers consist of interconnected aggregates of particles of various sizes which form large pore channels between the aggregates for the transport of the mobile phase. Among the aggregates or clusters there exist small pores for separations. The small particles are formed from tightly packed extremely small particles ca 100-300 nanometers.
The materials made in accordance with these patents have a disadvantage in that the micropores within or between these particles physically trap the sample molecules and degrade the separation. Although these patents claim that there are no interstitial spaces in the monolithic media as in the packed bed with beads, the large channels between the aggregates and interconnected particles actually cause the same problem as the interstitial spaces between the beads in conventional packed columns with beads. The large channels formed from various size of aggregates or clusters are inhomogeneous and provide random interstitial spaces, even with narrow particle size distribution. Because of the random interstitial spaces the column efficiency is poor.
U.S. Pat. Nos. 5,334,310, 5,453,185 and 5,728,457 disclose the preparation of the separation media inside a column with cross section area from square micrometers to square meters. The processes disclosed in these patents have some disadvantages. Some of the disadvantages were disclosed by the inventors named in those patents in 1997 in Chemistry of Materials, 1997, 9, 1898.
One significant disadvantage is that larger diameter (26 mm I.D.) columns prepared from the above patented process have a pore size distribution is too irregular to be effective in chromatography separation. The irregular pore size distribution is caused by the detrimental effect of polymerization exotherm, the heat isolating effect of the polymer, the inability of heat transfer, autoaccelerated decomposition of the initiator and concomitant rapid release of nitrogen by using azobisisobutyronitrile as initiator in a mold with 26 mm diameter. It has been found that the temperature increase and differential across the column created by the polymerization exotherm and heat transfer difficulties results in accelerated polymerization in large diameter molds such as for example molds having a diameter of more than 15 mm and in a temperature gradient between the center of the column and the exterior wall of the column which results in inhomogeneous pore structure. It was suggested in this article that the problem might be reduced by slow addition of polymerization mixture. This helps to solve the problem partly but does not solve the problem completely. There is still a temperature gradient for the larger diameter columns, which result in in-homogeneity of the pore size distribution.
This problem was also verified by theoretical calculations in the publication of Analytical Chemistry, 2000, 72, 5693. This author proposes a modular approach by stacking thin cylinders to construct large diameter columns for radial flow chromatography. However, sealing between the discs to form a continuous plug is difficult and time consuming.
U.S. Pat. Nos. 5,334,310, 5,453,185 and 5,728,457 disclose the material of weak anion exchange and reversed phase columns. The weak anion exchanger prepared had low resolution, low capacity, low rigidity, slow separation and very poor reproducibility. The reversed phase media has very little capacity, non-ideal resolution, and very poor reproducibility. They can not be used in mobile phase with high water content such as less than 8% acetonitrile in water due to wall channeling effect resulting from shrinkage of the very hydrophobic media in this very polar mobile phase. This media is also compressed during separation and result in excess void volume in the head of the column. The above patents provide little guidance on how to prepare a weak cation exchanger, strong cation exchanger, strong anion exchanger, normal phase media and hydrophobic interaction media. These media based on membrane, beads or gels are known. However, the preparation are done by off-line and can not be used for in situ preparation of monolithic columns. The monolithic membrane prepared according to U.S. Pat. Nos. 2,889,632, 4,923,610 and 4,952,349 has low capacity and resolution.